Category
8,786 papers
It is argued that behavioral evaluation understates what models encode, while probing alone overstates what they deploy, in a three-level evaluation framework measured on the same items under the same binary decision.
We present ChunkRank, an open-source Python library that derives chunk boundaries from a target model's tokenizer and context window, and selects an answer among candidates produced independently per chunk. It ships a validated registry of 90 models across 15 providers and six answer-selection methods, and needs only t...
Results show that task-informed subspace construction can be effective for low-resource speech adaptation, and that adaptive rank allocation provides a robust way to improve parameter efficiency without increasing model capacity.
It is suggested that a cascade of judges succeeds only when its judges make complementary errors, and that future decision models should be designed afresh with that aim.
Reach audiences
Advertise in front of researchers, engineers, and readers.
AI-workflow governance cannot be reduced to checking the final answer: an apparently safe answer may rest on a fragile evidence path that ordinary evaluation cannot see, localize or govern, so SSEG moves governance below surface-level output checking into auditable, path-specific decisions about intervention, revalidat...
The effectiveness of hybrid text privatization is demonstrated, which leverages the promise of DP in an efficient, usable manner and effectively reduces re-identification risks in clinical texts while being producing more coherent and usable output texts, leading to higher privacy-utility trade-offs.
A black-box, inference-time diagnostic that tells these two cases apart without retraining or annotation is introduced, with the pattern holding on English-Marathi and English-Tamil, with the failure modes tracking the post-edit distribution rather than MT quality or language family.
LLMersion is proposed, a scheme for AI for education that runs entirely at home, over the learner's own documents, with an AI-written, AI-understood, AI-updated codebase anyone can customize; and presents LLMersion-1, a released open-source prototype.
This paper applies speech act and politeness theory to a corpus-pragmatic analysis of 2,000 English-language prompts drawn from publicly shared ChatGPT conversations, showing a consistent movement toward indirect, implicit, and fragmentary realizations of directive force, accompanied by a decline in politeness marking.
Results show that benchmark accuracy does not necessarily imply execution-safe SQL, and that lightweight, multiplicity-aware runtime guardrails can narrow this gap with modest computational overhead, and introduces ModularSQL, a lightweight post-selection runtime guardrail.
It is demonstrated that fine-tuned NMT significantly outperforms few-shot LLMs for Arabic-Russian translation under low-resource conditions.
This work introduces StepCOPS, which uses an independent proposal split to nominate one lower-tail floor per candidate, exact binomial tests on a fresh certification split, and Holm's step-down procedure to certify a set of floors.
Flint-interactive gives AI agents a language for creating interactive visualizations. Click a chart, filter data, highlight related content, and explore beyond the view the author originally chose. The post Flint-interactive: A language for creating interactive data visualizations appeared first on Microsoft Research.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
A new computational approach reveals how cells communicate to assemble into tissues. This could accelerate stem cell engineering, improve disease models, and inform therapies.
Zero-knowledge proofs can verify computing results without rerunning workloads or exposing sensitive data. Magnetar helps bring them into the mainstream with support for standard programming languages, major speed gains, and quantum-resistant security. The post Proof-based general-purpose computing with Magnetar appeared first on Microsoft Research.
A new language-processing tool could help identify the highest-risk individuals from natural language, enabling swifter interventions.
A Blog post by Liquid AI on Hugging Face
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
